Encrypted login | home

Program Information

Metal Artifact Reduction with An Additional Tilted CT Scan: A Preliminary Study

no image available
C Kim

C Kim1*, J Kwak2 , C Lee2 , B Cho2 , C Jung2 , S Cho1 , S Lee2 , (1) KAIST, Daejeon, (2) Asan Medical Center, Seoul

Presentations

TU-C2-GePD-IT-3 (Tuesday, August 1, 2017) 10:00 AM - 10:30 AM Room: Imaging ePoster Theater


Purpose: To reduce metal-induced artifacts in CT images by replacing artifacts-contaminated CT data with complementary information from an additional tilted CT scan.

Methods: Metallic implants can introduce streak and shadow artifacts in CT images, which degrades the image quality severely and causes errors in dose calculation of radiotherapy planning.The basic idea of our proposed method is to acquire an additional tilted CT scan. As metal artifacts mainly traverse the object along beam direction in CT scan, such tilted CT scan would provide complementary information where some regions were free of metal artifacts. Then, we selected the pixels which contain fewer metal artifacts in the two images to build a combined image with the best of the information. Moreover, we introduced another metal artifact reduction method based on image mutual correlation, as image mutual correlation can be used to represent the degree of artifact contamination and to steer data selection. To validate the proposed method, we performed a simulation study with the pelvic part of the XCAT numerical phantom. We acquired artifact-free data and added metal artifacts by inserting metallic implants in the numerical phantom. Then we compare the performance of the proposed method with the existing metal artifact reduction method.

Results: Quantitative assessment of the percentage average total deviation on soft tissue revealed that the proposed method decreases the percentage errors within 2%.

Conclusion: We have demonstrated the feasibility of proposed metal artifact reduction method with an additional tilted CT scan. Without burdensome calculations, the proposed method was comparable to the existing algorithms by using complementary information as much as possible. Future works will include experimental studies to find and verify clinical usefulness.

Funding Support, Disclosures, and Conflict of Interest: This research was supported by the National Research Foundation of Korea (NRF) funded by the Korea government (MSIP) (2014R1A1A2058154 and 2015M2A2A6A02045253)."


Contact Email: